Existing treatment approaches for progressive diabetic nephropathy include anti-hypertensive agents such as angiotensin converting enzyme inhibitors (ACEi) and angiotensin II receptor blockers (ARBs). Although recent studies demonstrate the utility of this approach in both type 1 and type 2 diabetic patients, the potency of these agents to arrest diabetic nephropathy in patients with declining renal function is not clear. Although, further modifications of the renin-Angll-aldosterone system may prove to be beneficial, high doses of these agents are difficult to use in patients with declining renal function due to hyperkalemia, low blood pressure, and drug-induced reduction of renal perfusion. The recognition that progressive diabetic nephropathy is largely due to progressive glomerulosclerosis and tubulointerstitial fibrosis has focused attention on mediators and blockers of this process. Our prior studies have helped to establish that a key factor in the mediation of progressive matrix deposition in the diabetic kidney is transforming growth factor-alpha (TGF-a). However, direct inhibitors of the TGF-a system are not yet available for clinical trials. We have recently identified that an orally administered drug, pirfenidone, is extremely potent to block progressive glomerulosclerosis when administered after the onset of disease in a mouse model of diabetic nephropathy. Pilot data in patients with type 1 diabetes and nephropathy indicate that this drug is well tolerated. In addition, preliminary data from a recent open-label study in patients with advanced focal scelrosis suggests that pirfenidone may lower the rate of decline of renal function. However, the mode of action of pirfenidone is unclear. Based on the encouraging data in animal models and small clinical studies, we propose to undertake a trial to evaluate the role of pirfenidone in patients with type 1 and type 2 diabetes and renal insufficiency. We propose a randomized, double-blind, placebo study to assess whether pirfenidone will lower the rate of progression of diabetic nephropathy. In addition, the clinical study will determine whether serial measures of TGF-a may be used as a biomarker to identify patients who may have a rapid decline in renal function and to determine if the therapeutic benefit of pirfenidone is indeed via inhibition of the TGF-a system.